Humanoid Muscle-Skeleton Robot Arm Design and Control Based on Reinforcement Learning

Jianyin Fan; Jing Jin; Qiang Wang

doi:10.1109/ICIEA48937.2020.9248350

Humanoid Muscle-Skeleton Robot Arm Design and Control Based on Reinforcement Learning

Jianyin Fan
, Jing Jin
, Qiang Wang

School of Astronautics

Harbin Institute of Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

9 Scopus citations

Abstract

Muscle-skeleton robots share similar appearances and functions with humans, making these robots more adaptive in human interaction scenarios. In this paper, a new muscle-skeleton robot arm driven by artificial muscles is proposed. First, we design a new multifilament McKibben muscle and measure its properties. Then a humanoid robot arm referred to the anatomy of the human arm is presented, while the configuration of muscle is adjusted to reduce the complexity of manufacturing and controlling. Muscle-skeleton robot arms with different muscle configurations are controlled using the reinforcement learning method in the simulation environment, and different arm models' movement ranges are obtained to find the best muscle configuration. The experimental results show that the model with the best muscle configuration achieves 79.8% of the whole movement range.

Original language	English
Title of host publication	Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	541-546
Number of pages	6
ISBN (Electronic)	9781728151694
DOIs	https://doi.org/10.1109/ICIEA48937.2020.9248350
State	Published - 9 Nov 2020
Event	15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020 - Virtual, Kristiansand, Norway Duration: 9 Nov 2020 → 13 Nov 2020

Publication series

Name	Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020

Conference

Conference	15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020
Country/Territory	Norway
City	Virtual, Kristiansand
Period	9/11/20 → 13/11/20

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Keywords

humanoid robot arm
multifilament McKibben muscle
reinforcement learning

Access to Document

10.1109/ICIEA48937.2020.9248350

Cite this

Fan, J., Jin, J., & Wang, Q. (2020). Humanoid Muscle-Skeleton Robot Arm Design and Control Based on Reinforcement Learning. In Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020 (pp. 541-546). Article 9248350 (Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICIEA48937.2020.9248350

Fan, Jianyin ; Jin, Jing ; Wang, Qiang. / Humanoid Muscle-Skeleton Robot Arm Design and Control Based on Reinforcement Learning. Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020. Institute of Electrical and Electronics Engineers Inc., 2020. pp. 541-546 (Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020).

@inproceedings{78f159625bea493eaccd21616e24816e,

title = "Humanoid Muscle-Skeleton Robot Arm Design and Control Based on Reinforcement Learning",

abstract = "Muscle-skeleton robots share similar appearances and functions with humans, making these robots more adaptive in human interaction scenarios. In this paper, a new muscle-skeleton robot arm driven by artificial muscles is proposed. First, we design a new multifilament McKibben muscle and measure its properties. Then a humanoid robot arm referred to the anatomy of the human arm is presented, while the configuration of muscle is adjusted to reduce the complexity of manufacturing and controlling. Muscle-skeleton robot arms with different muscle configurations are controlled using the reinforcement learning method in the simulation environment, and different arm models' movement ranges are obtained to find the best muscle configuration. The experimental results show that the model with the best muscle configuration achieves 79.8\% of the whole movement range.",

keywords = "humanoid robot arm, multifilament McKibben muscle, reinforcement learning",

author = "Jianyin Fan and Jing Jin and Qiang Wang",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020 ; Conference date: 09-11-2020 Through 13-11-2020",

year = "2020",

month = nov,

day = "9",

doi = "10.1109/ICIEA48937.2020.9248350",

language = "英语",

series = "Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "541--546",

booktitle = "Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020",

address = "美国",

}

Fan, J, Jin, J & Wang, Q 2020, Humanoid Muscle-Skeleton Robot Arm Design and Control Based on Reinforcement Learning. in Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020., 9248350, Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020, Institute of Electrical and Electronics Engineers Inc., pp. 541-546, 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020, Virtual, Kristiansand, Norway, 9/11/20. https://doi.org/10.1109/ICIEA48937.2020.9248350

Humanoid Muscle-Skeleton Robot Arm Design and Control Based on Reinforcement Learning. / Fan, Jianyin; Jin, Jing ; Wang, Qiang.
Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020. Institute of Electrical and Electronics Engineers Inc., 2020. p. 541-546 9248350 (Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Humanoid Muscle-Skeleton Robot Arm Design and Control Based on Reinforcement Learning

AU - Fan, Jianyin

AU - Jin, Jing

AU - Wang, Qiang

PY - 2020/11/9

Y1 - 2020/11/9

N2 - Muscle-skeleton robots share similar appearances and functions with humans, making these robots more adaptive in human interaction scenarios. In this paper, a new muscle-skeleton robot arm driven by artificial muscles is proposed. First, we design a new multifilament McKibben muscle and measure its properties. Then a humanoid robot arm referred to the anatomy of the human arm is presented, while the configuration of muscle is adjusted to reduce the complexity of manufacturing and controlling. Muscle-skeleton robot arms with different muscle configurations are controlled using the reinforcement learning method in the simulation environment, and different arm models' movement ranges are obtained to find the best muscle configuration. The experimental results show that the model with the best muscle configuration achieves 79.8% of the whole movement range.

AB - Muscle-skeleton robots share similar appearances and functions with humans, making these robots more adaptive in human interaction scenarios. In this paper, a new muscle-skeleton robot arm driven by artificial muscles is proposed. First, we design a new multifilament McKibben muscle and measure its properties. Then a humanoid robot arm referred to the anatomy of the human arm is presented, while the configuration of muscle is adjusted to reduce the complexity of manufacturing and controlling. Muscle-skeleton robot arms with different muscle configurations are controlled using the reinforcement learning method in the simulation environment, and different arm models' movement ranges are obtained to find the best muscle configuration. The experimental results show that the model with the best muscle configuration achieves 79.8% of the whole movement range.

KW - humanoid robot arm

KW - multifilament McKibben muscle

KW - reinforcement learning

UR - https://www.scopus.com/pages/publications/85097562195

U2 - 10.1109/ICIEA48937.2020.9248350

DO - 10.1109/ICIEA48937.2020.9248350

M3 - 会议稿件

AN - SCOPUS:85097562195

T3 - Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020

SP - 541

EP - 546

BT - Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020

Y2 - 9 November 2020 through 13 November 2020

ER -

Fan J, Jin J , Wang Q. Humanoid Muscle-Skeleton Robot Arm Design and Control Based on Reinforcement Learning. In Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020. Institute of Electrical and Electronics Engineers Inc. 2020. p. 541-546. 9248350. (Proceedings of the 15th IEEE Conference on Industrial Electronics and Applications, ICIEA 2020). doi: 10.1109/ICIEA48937.2020.9248350

Humanoid Muscle-Skeleton Robot Arm Design and Control Based on Reinforcement Learning

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this